Stored program controlled communication switching systems comprise some form of intelligence which controls the switching functions in response to a program stored in memory. Historically, such systems included a single processing entity for the control of the entire system. As technology and system design evolved, it was found desirable to separate certain routine functions from the main processing entity to save its processing time for more complex system functions and decisions. Systems, called distributed control systems, are presently being designed which also separate some of the more complex system functions and decisions into several intelligent processors, each of which controls associated switching system functions.
Stored program controlled switching systems include data storage areas for storing the data necessary for the completion of calls. In a distributed control system, a given call may involve more than one processor, which processors may require access to the same call related data. When all call related data is not locally associated with each processor, system level messages must be exchanged between a given processor and the storage areas storing the call related data, to locate the retrieve that data. However, when all call related data is stored in local association with each processor, called data replication, each processor has rapid access to that data without system level messages, but the total amount of data stored is much greater than if it is stored in a single location. In fact, if n processors are used, data replication may result in n times the single storage area.
Applicants propose an arrangement in which each processor is locally associated with the data relating to its primary function. Advantageous results are obtained when this data distribution arrangement is coupled with applicants' arrangement for communicating call related data and control information among the distributed processors of the system.